Talk:Djinni

I wonder why they didn't just call these guys "efreets" since that's basically what they are. Not that that will stop me from trying at least a few of them when 0.13 comes out. :) --spudwalt 21:43, 2 June 2013 (CEST)

Efreets (correct plural is actually afarit) are basically just a race of evil jinn. So it makes more sense to go with the more general (d)jinni. Of course, dropping the initial d would also help, since that's a fairly archaic transliteration convention. -Ion frigate 02:29, 3 June 2013 (CEST)

Are these permanent?

Just a quick question: is it already decided that the new species (and other features like the new monsters) will be in the release version of 0.13, or are they implemented for testing? I'm pretty new to the updating process, as I previously ignored trunk because of the warning that it might be 'unstable'. --Lokkij 22:39, 2 June 2013 (CEST)

They have been implemented for testing, and I think the idea is if there is no serious problems they will keep them in for the next release. Trunk is still unstable in the sense that there might be bugs related to the species, as well as balance changes that wouldn't occur in a release version. --Flun 03:44, 3 June 2013 (CEST)

Help explaining the Djinn magic aptitude

I wrote the following:

---

Also due to how XP costs work, Djinn can be much better than a "-3 aptitude" would indicate.

XP cost rises as level rises. This has an odd effect when considering Djinn's casting aptitude. Take, for example, a single-school Conjurations spell:

• A Kobold has +0 aptitudes in Spellcasting and Conjurations. If you only cared about spell power/failure, a Kobold would be most efficient a point like {5 Spellcasting, 15 Conjurations}. This would cost 72.75 skill points.
• A Djinni would reach the same power as the Kobold at {13 Spellcasting}, all else equal. This would cost 80.72 skill points, which is only +11.7% more XP than the Kobold. That's an effective aptitude of -0.6, much better than -3 appears.
• Note that if the Kobold were to train more Spellcasting, for either MP or spell level purposes, they'd become less efficient. A Kobold at {10 Spellcasting, 13.5 Conjurations} would get the same power, but at a cost of 79.13 skill points.

For dual-school spells, Djinn only become better. For casting Fireball, a Kobold would be efficient at {8 Spellcasting, 15 Fire, 15 Conj}, costing 153 skill points. A Djinni equals that at {13.6 Spellcasting}, or 89.05 skill points, which is 58% the XP. That's an effective aptitude of +3.

The exact efficiency depends on the level of skill. At low skill levels, Djinn are truly around a -3 aptitude for single-school spells. At XL 27, Djinn can be cheaper than Kobolds, even for single-school spells.

---

I wonder if it could be explained better? Hordes (talk) 18:37, 8 November 2023 (CET)

Good stuff! Although, skill points should be "7275" instead of "72.75", etc. in this example. The wiz-mode version of the skills menu shows them w/o a decimal point. Also, from Manual: "Each manual generates with 2000-3000 skill points of XP." Ge0ff (talk) 00:39, 10 November 2023 (CET)

I think we should use the 1/100 scale for skill points, because that is how it's displayed in game outside wizmode (the "Cost" section), and what the skill point article uses. I think it's better to change manual / Ru / (other pages) with the 1/100 scale.

I also think that "XP cost rises as level rises" doesn't make a lot of sense, unless you already know about the topic. For me, "To maximize power, Kobolds would train to 15 Conjurations. Djinn only need 13 Spellcasting, so their level is lower, so their XP cost is lower" makes sense. But it might not for the average reader. Hordes (talk) 02:10, 10 November 2023 (CET)

"that is how it's displayed in game outside wizmode". Nope:

• Skill points are stored in an array of integers, you.skill_points[]. If a Dj has 13 Spellcasting, then you.skill_points[SK_SPELLCASTING] is equal to 8072, not 80.72.
• The Cost section of the skills menu is not measured in skill points. It's a ratio, one_level_cost() / skill_cost_baseline(), where the numerator and denominator have the same unit of measurement, XP (see scaled_skill_cost() in skills.cc).
• In Skill_point#Specific_mechanics, the "Skill point cost" column is not measured in skill points either. Also, the numbers in the table and the graph are quite outdated (since 0.28-a0-510-g2e89a317aa). E.g., rising a +0 apt skill from level 23 to 24 has cost = 34.5; for a -1 skill it's 41.0 now, but definitely not 260+, as in the table. Ge0ff (talk) 14:31, 10 November 2023 (CET)

"I also think that "XP cost rises as level rises" doesn't make a lot of sense, unless you already know about the topic." No, it does make sense, even if you don't know much about the skills system. It requires only a minimal observation skill from the player/reader. It's easy to spot the difference between drinking a potion of xp at xl 1 and xl 27, or between killing Sigmund at xl 2 vs. xl 27. Ge0ff (talk) 14:42, 10 November 2023 (CET)

RE: XP Cost statement. It's not that "XP cost rises as level rises" doesn't make sense as a statement. It's that it would be hard for a new reader to understand why this is relevant, i.e. why this fact makes Djinn better. In the current example, it states that Kobolds are most efficient at 15 Conjurations, but Djinn are equivalent at 13 Spc. Thus the correlation is implied (15 -> 16 more expensive than 13 -> 14) but not explicitly stated. Hordes (talk) 16:38, 10 November 2023 (CET)

If you think that adding more details would improve the quality of the article, then absolutely go for it. If an explanation gets too wordy, we could be trim it later. Ge0ff (talk) 21:52, 10 November 2023 (CET)